Differences

This shows you the differences between two versions of the page.

--- aira:start [2025/10/16 09:56] – [Schedule Autumn 2025] mzk
+++ aira:start [2025/10/16 10:00] (current) – [2025-06-26] mzk
@@ Line 456: / Line 456: @@
+==== 2025-10-16 ====
+<WRAP column 15%>
+{{ :aira:karol-dobiczek-foto.jpg?200| }}
+</WRAP>
+<WRAP column 75%>
+**Speaker**: Karol Dobiczek,  PhD Candidate @ Jagiellonian University
+**Title**: Applying Counterfactual Explanations in Evolving Scenarios and Expert Domains.
+**Abstract**:
+Counterfactual explanations (CE) are one of the building blocks of many explainable machine learning methods, however their application and effects when applied to realistic scenarios are often missing. This seminar presents two works on counterfactual explanations that address some of these scenarios. First work explores the effects of applying CEs in an evolving domain and model, the second investigates whether existing methods for natural language CEs can handle being applied to expert domains. Other current and future directions in applying CEs and other XAI methods will be discussed.
+**Biogram**:
+Karol Dobiczek is a PhD candidate in the team led by professor Grzegorz J. Nalepa at the Jagiellonian University in Kraków. He received his Master's degree from the Delft University of Technology in 2024 and his Bachelor's thesis from the same university in 2022. His work focuses on explainable and interpretable machine learning, mostly on the use of counterfactual explanations. Karol is currently a Software Engineer at Qualtrics and a Research Software Engineer at the MI2 team at the Warsaw University of Technology.
+</WRAP>
+<WRAP clear></WRAP>
 ==== 2025-06-26 ====